help for agentless port monitoring

[karaul01]

i have a problem with simple check on zabbix 3.0 .
i can use telnet to destination ip port

telnet 10.4.3.17 2002
Trying 10.4.3.17...
Connected to 10.4.3.17.

my configuration is below. But always seen down in monitoring tab. i tried different parameters for service; such as net.tcp.service[telnet , net.tcp.service[http .. Please help me on this issue



trigger expression : {10.4.3.17:net.tcp.service[tcp,,2002].last()}=0

[Linwood]

The first parameter does change the check in some cases, see here from the manual.

I found for example I could not test most network devices with "https" because of recent changes to the libcurl utilities that won't do SSLV3 that is still required on many devices internally, so had to use "tcp" as the first parameter for port 443. It documents telnet as requiring a ":" at the end of the prompt (overly restrictive I think in some cases).

You should different checks, I would have expected the one in the trigger to work (just "tcp"), does it not?

If that's not it, you may need to turn on debug while it runs, and see what it is getting. Alternatively (and perhaps easier) is wireshark or tcpdump and make sure you are connecting as expected, and something else isn't happening differently when zabbix tries.

[karaul01]

Thanks a lot Linwood . After your answer, I changed ip port which i want to monitor. The new iport belongs to zabbix server. Moreover, i try different parameters like telnet, tcp, ftp..
When i enable the item, i cant see network activity via netstat . And it has still seen down.

I wonder is there a wrong configuration, because zabbix show down(0), different port of its own server.

Thanks again for suggestions

[karaul01]

this commands is solution. Thank you again

setsebool -P httpd_can_network_connect 1
setsebool -P zabbix_can_network 1